Applying principles of agonistic politics to institutional design

Transcription

1 Applying principles of agonistic politics to institutional design Manon Westphal - DRAFT- 1 Introduction Agonism has become known as a distinct current in democratic theory above all because of its thorough critique of political liberalism and deliberative democracy. According to agonists, the potential for conflict that results from pluralism embraces not only ideas of the good but also ideas of the right. [A]gonistic approaches typically insist that the democratic contest can and should go all the way down to include the principles and procedures that are supposed to regulate political life. (Schaap 2009: 1) This assumption undermines the crucial guideline of the theories of John Rawls, Jürgen Habermas and others in their traditions, namely the notion of public reasoning according to which political actors must invoke reasons that everybody can accept in order to make sure collectively binding decisions are justifiable to all. In arguing that reasons of this kind do not exist, agonists claim the need for a substantially different approach to politics. Whereas it is indisputable that the agonistic critique implies this call for a substantial alternative, several commentators question that theories of agonistic democracy developed by theorists such as William Connolly, Bonnie Honig, Chantal Mouffe and James Tully actually live up to this goal. Critics confront agonists with two severe objections. The first is that their theories rely on normative presuppositions that resemble those of their liberal and deliberative counterparts. The second is that agonistic democracy suffers from an institutional deficit. As agonists do not explain what their argument for an acceptance of deep dissent means with respect to the organisation of political institutions and procedures, the critics contend, agonism leaves democracies without practical guidance. In this paper, I probe on agonism s potential to provide guidance for institutional design. Like many other observers, I consider the agonistic critique of public reason-centred political theory convincing and agonists lack of engagement with political institutions unsatisfactory. Pluralist democracies are in need of practical solutions to the challenge of pluralism and in order to support their search for solutions, agonists should translate their insights about the nature of this challenge into practical advice. But elaborating on agonism s institutional implications does not only mean to sound out options for rebutting the claim that agonistic democracy s current institutional deficit indicates a weakness of the agonistic perspective itself. It also enables an assessment of the first objection mentioned. The challenge this objection confronts agonists with is not just to prove that agonism differs from political liberalism and deliberative democracy I think this is hard to deny but to prove that the differences are so significant that they 1

2 validate agonists claim to offer meaningful alternatives. By rendering visible the institutional implications of agonism, we provide a sound measure that can be used to assess the weight of the differences. Because if the conceptual differences among agonistic democracy and public reason-oriented political theory result into diverging recommendations for institutional design, the claim that these differences are irrelevant for practice can hardly be maintained. Thus, closer elaborations on agonism s potential to provide guidance for institutional design are desirable as they promise to both render visible the practical relevance of agonism and sharpen our understanding of agonism as a distinct theoretical approach. With this paper, I aim to contribute to a deepening of political theory s engagement with the institutional side of agonistic democracy by proceeding as follows. In a first step (2), I suggest an approach that enables the application of agonism to institutional design in a way consistent with the basic premises of agonism. In a second step (3), I deal with the accounts of agonistic democracy that have been developed by Mouffe, Connolly and Tully. I sketch these accounts respective visions of agonistic politics and put emphasis on their normative nature by exposing the principles they suggest should guide democracy s dealings with pluralism. In a third step (4), I apply these principles to institutional design. I use an institution that has been increasingly attracting the attention of deliberative democrats, mini-publics, as an example to discuss how agonism changes our view on the political value of institutions. First, I sketch the reasons that lead deliberative democrats to appreciate mini-publics as particularly worthwhile democratic innovations and track how these reasons lead to the endorsement of specific design features (4.1). Second, I discuss whether and in which ways mini-publics can be conducive to agonistic politics and identify features of institutional design that could help mini-publics to unfold their agonistic potential (4.2). 2 Agonism and institutional design An approach A discontent with agonists lack of engagement with the formal institutions and procedures of political decision-making is not only expressed by critics of agonism (e.g., Deveaux 1999: 14; Dryzek 2005: 221f) but also by theorists who affirm the basic notions of agonism. In contrast to the former, the latter do not interpret this lack of engagement as the expression of a genuine weakness but as a remaining task for agonists to realise: [T]hough these accounts [the agonistic accounts, M. W.] allude to the importance of democratic rules and procedures, there is still something of an institutional deficit in their respective theories, both in terms of their critiques of existing arrangements and in terms of their more positive alternatives. More detailed work is thus required to elaborate the institutional, procedural and spatial underpinnings of a radical and plural Democracy [ ]. (Howarth 2008: 189) 2

3 Similarly, Thomas Fossen claims agonism needs to engage questions of law and institutions more thoroughly in order to distinguish itself as a mature current of its own, rather than a footnote to liberal and deliberative accounts of politics (Fossen 2012: 331). As indicated in the introduction, I support the idea that through overcoming its institutional deficit agonism would strengthen its status as a distinct current in democratic theory. But whereas voices demanding an overcoming of this deficit abound, answers to the question how it can be realised are rare. The approach I want to introduce in this paper builds on two considerations. Both are outcomes of a critical engagement with the current state of affairs in agonistic theory discourse whose characteristics are no minor reasons for why, first, so few attempts to overcome the institutional deficit have been undertaken and, second, the only study that systematically elaborates on the institutional side of agonistic democracy suffers from a serious weakness. On the one hand, there is a tendency to understand the nature of agonism as political ethos. This understanding is problematic because it overstretches the relevance of one of agonism s elements. On the other hand, there is a tendency to conceive of the task to overcome agonism s institutional deficit as amounting to the challenge to identify genuinely agonistic institutions, institutions that are most conducive to agonism. Whereas the latter tendency is not a widespread one simply because there is no larger number of studies dealing with the issue available the only study that systematically addresses agonism s institutional deficit is expressive of exactly this idea. I consider both tendencies to be in need of revision. Not only is it possible to develop more convincing notions of both the nature of agonism and the relationship between agonistic theory and institutional design; they actually present obstacles to fruitful applications of agonism s insights to institutional design. In the rest of this section, I deal with both tendencies in more detail and suggest alternative interpretations against the background of which the undertaking to overcome agonism s institutional deficit becomes a manageable and promising one. 2.1 The nature of agonism Let us first address the challenge of grasping the nature of agonism. Surely, it is far from easy to define the nature of a strand of political theory which, given the diversity of agonistic theories, is constituted by a range of perspectives joint by family resemblances rather than one shared conceptual core. Nevertheless, attempts to identify agonism s institutional implications require a notion of what agonism is. The critique of political liberalism and deliberative democracy does not suffice as a basis for conclusions about the organisation of institutions and procedures. The idea that pluralism and dissent embrace even the fundamentals of justice and thus render hopeless any search for a moral consensus does not point to what agonism is and demands but what it denies and rejects and, as such, cannot serve as a measure to evaluate institutional design. So what is it that agonism affirms? Most commentators argue that agonism is the 3

4 affirmation of political contestation and struggle (e.g., Wenman 2003). And as all the theories usually subsumed under the term agonistic democracy advocate the idea that conflicts should be acknowledged as the primary modes of political interaction, it indeed seems appropriate to define agonism as a strand of democratic theory whose central characteristic is the affirmation of contestation and struggle. This definition points to the (constructive) reverse side of agonists (deconstructive) charge against public reason-oriented political theory: If harmonious solutions to disagreements are not available, they need to be carried out in non-harmonious ways. But this is still a very general definition. Theorists who intend to probe on agonism s institutional implications face the challenge to become more concrete as it seems impossible to present conclusions about agonistic institutional design without a more detailed understanding of how, according to agonists, this affirmation of contestation is realised. There is one particular group of scholars whose more detailed definitions of agonism we are well advised to consider. This is proponents of agonism who suggest a rather sceptical stance towards undertakings targeted at an overcoming of the institutional deficit because they, like the critics, take the lack of engagements with institutions to be a natural implication of agonistic theory. It is obvious why it is crucial to consider what, according these theorists, constitutes the nature of agonism. If they have detected features in agonistic understandings of how the affirmation of contestation and struggle is realised, all those claiming the need to overcome agonism s institutional deficit should rethink their position. An application of agonistic theory to institutional design only makes sense for the status of agonistic democracy as a distinct strand of political theory as well as for democracies whose practice proposals for institutional design are supposed to improve if it makes visible its strength. In case there is something to agonism that makes the very purpose of applying it to institutional design run counter to its basic features, it will be wiser to dismiss this purpose. The idea that institutions are of secondary importance for agonistic democracy is articulated, for instance, by Paulina Tambakaki and Andrew Schaap (Tambakaki 2011, Schaap 2006) who both point to the same feature of agonism when explaining why institutions play a minor role for its realisation. According to Tambakaki, agonistic democracy requires that citizens develop a critical ethos or attitude towards [.] politics and begin to resist, disturb and contest that which appears natural, hegemonic or final be it rules, narratives, directorates or [.] policies. (Tambakaki 2011: 575) In line with this idea, Schaap states that agonism s mistrust towards institutions results from the fact that ongoing contestation requires an ethos that drives citizens to continuously challenge moments of closure. This is because they [the agonists, M. W.] seek to understand democracy not, primarily, in institutional terms but as an ethos that seeks to postpone the moment of decision in order to affirm the openness of political life. (Schaap 2006: 270) Far from considering the agonistic understanding of democracy as ethos (ibid.) a 4

5 weakness, Schaap stresses that it provides [.] conceptual resources in terms of which to conceptualize reconciliation as a political undertaking that resists cooptation by modern states (ibid.); Tambakaki argues that the absence of an institutional design is perhaps one of the greatest merits of the agonistic vocabulary (Tambakaki 2011: 578). Against the background of the above mentioned considerations it should be clear that I doubt agonists lack of engagement with institutions presents a merit. But Schaap and Tambakaki surely have a point when suggesting that an ethos of disturbance can hardly find expression in formalised settings. Whereas it seems difficult to institutionalise any kind of ethos because the very term signifies an attitude rather than a rule, this applies even more in the case of an ethos of disturbance that aims to continuously unsettle existing rules. That means, if agonists argument for prioritising contestation and struggle amounts to an understanding of politics as an ethos of disturbance, the undertaking to overcome the institutional of agonistic democracy is more or less doomed to fail. The crucial point, though, is that this definition of agonism is not fully convincing. It builds on one aspect that undoubtedly plays an important role in all accounts of agonistic democracy but does not suffice for a description of their nature. In short, agonistic democracy does not run out in the realisation of a political ethos. What characterises the nature of agonism beyond its basic understanding of politics as contestation and struggle is a decisive understanding of how struggles should be carried out. That is, every agonistic theory gives an account of the process through which the conflicts regularly resulting from citizens dissent over even of basic norms should be dealt with. Whereas each agonistic theory aims at enabling cooperation under conditions of enduring dissent, their respective understandings of how this aim is to be realised differ. My claim is that in the face of these more detailed conceptions of adequate political responses to contestations, agonistic theories reveal their nature as theories of conflict regulation. This definition of agonism allows for projects that aim at an overcoming of the institutional deficit. Because if agonism does not run out in a particular notion of political ethos but constitutes particular notions of politics dealing with conflicts, the question that needs to be addressed in relating agonism to institutional practice would no longer be How can this ethos be realised but turns into How can this kind of conflict regulation be realised. And in contrast to the former, the latter allows for institutions to play a fruitful role. Institutions can be designed in different ways that either hinder or promote specific modes of conflict regulation, and as soon as we have a clear idea of what features of agonistic modes of conflict regulation are, we also have measures that can be applied for an evaluation of the agonistic value of institutions and procedures. To prove this idea, I deal with Mouffe s, Connolly s and Tully s theories of agonistic democracy in section three. I will trace their main lines of argumentation, lay out their respective accounts of adequate dealings with conflicts, and show that their notions of ethos owe their importance 5

6 to their enabling such processes of conflict regulation. Before addressing these issues, though, we tackle the second question that must be answered before attempts to overcome agonism s institutional deficit can be undertaken: What does the task of identifying agonism s institutional characteristics amount to? What kind of results should political theorists aim at when applying agonism to institutional design? 2.2 Agonism and institutional design A fruitful starting point to answer these questions is the consideration of an objection that could still be raised against the idea of applying agonism to institutional design. This objection runs as follows. Even if agonism is expressive of specific perspectives on adequate ways of dealing with conflicts which, in principle, could be applied to institutional design, it is the very logic of setting up formalised rules that runs the risk of making agonism lose its radical bite. Agonism is based on the recognition of dissent and the related claim that, due to this dissent s depth, democracy should accept regularly emerging conflicts even about basic norms. Is it not, in the face of this plea for an affirmation of conflicts about potentially every subject, contradictory to suggest that it is possible to detect particular institutions and procedures that have a special status for agonism and, therefore, should not belong the range of contested issues? As long as agonists hold on to the idea that not even basic rights or the terms of discourse are beyond dispute, the critic might argue, they should also assume that, like any other issues, the question of what proper institutions look like will remain among the many subjects of dispute. I think this objection is an important one. But, rather than thwarting the project of elaborating on agonism s institutional implications altogether, it is helpful as it leads the way to an approach that manages to avoid the pitfalls it rightfully points to. An approach that is indeed thwarted by this objection is one that aims at a set of institutions that ideally realises agonistic politics. I call this the institutionalising agonistic democracy approach. The problem with this approach is that it seems to be caught in a trap of either fulfilling what the above mentioned objection alerts us to or presenting results that lack insights about the distinct character of agonism s implications for institutional design. And this is why. If theorists who aim at identifying the characteristics of an institutionalised agonistic democracy arrive at a set of institutions spelled out in detail by sketching, for example, the types of institutions as well as the relationships among them that realise agonistic politics they will present highly contestable results. It is more than likely that, depending on social and political contexts, very different institutional settings will prove most conducive to agonistic politics. Societal circumstances and the characteristics of pivotal conflict constellations will have an impact on the details of an institutional setting that constitutes the ideal stage for agonistic politics. A possible way to avoid this effect would be to dismiss the aim of spelling out in detail the institutional characteristics of agonistic democracy. Theorists 6

7 following this strategy will still present results about the institutional preconditions of agonistic politics but articulate them on a level of abstraction that leaves room for different specifications. To my view, the only systematic approach to agonism s institutional characteristics existing so far, Ed Wingenbach s Institutionalizing Agonistic Democracy (2011), is expressive of exactly this latter strategy. By taking a look at its results, we come to see the problems of this strategy. Wingenbach deals with three types of institutions participatory institutions, representative institutions and the institutions of Rawls s political liberalism and discusses whether these types have conducive effects for agonistic politics. He arrives at the conclusion that each of them furthers agonistic politics under certain conditions and, consequently, should be parts of an agonistic institutional setting. For Wingenbach, this implies that agonism explicitly situates itself within existing institutional forms, not outside them (Wingenbach 2011: 192) and, what is more, that its compatibility with institutions that are central for political liberalism proves that liberalism in its Rawlsian variant represents the best path for agonism (Wingenbach 2011: 193). Whereas I find the first part of Wingenbach s conclusions as convincing as unsurprising, the second simply does not follow. It would have been a surprising result that those institutions subject to Wingenbach s analysis could not play a role for agonistic politics whatsoever. But, notwithstanding the fact that this does not justify assessing these institutions as necessary or the most important institutions for agonistic democracy, one cannot conclude from this that the agonistic view merges with the liberal one. There are different ways to implement participation, representation or rights, and it can be the case that strongly diverging ways of implementation prove conducive to political liberalism but not to agonism or the other way around. Wingenbach thus glosses over variations that may emerge in a spectrum between outright rejection and affirmation. But if we drop the second part of his conclusions and are left with the insight that institutions important for liberalism can play a fruitful role for agonism too, we have hardly learnt anything about the institutional characteristics of agonistic democracy. And as soon as we become more specific by identifying the kinds of participatory or representative institutions that ideally help to realise agonistic politics, we move into the direction of the first strategy that is bound to face the objection of confining agonism to an institutional setting that will prove less conducive under different circumstances. This is why attempts to institutionalise agonism are caught in the trap of presenting results that are either too concrete and forfeit their claim of presenting the institutions of agonistic democracy, or too general and miss to tell us anything insightful about the institutional features of agonistic democracy. But the crucial point is that dismissing the idea to define agonistic democracy s ideal institutions does not mean dismissing the project of elaborating on agonisms implications for institutional design altogether. Rather, it means dismissing one specific interpretation of what probing on agonism s institutional implications could mean. I suggest an alternative approach able to avoid the above mentioned pitfalls. Its main idea is to replace the aim of identifying institutions that 7

8 ideally match agonism with the aim of using agonism as a critical measure for evaluations of particular institutions and design features. As argued above, agonistic theories display specific visions of what politics should look like in order to be responsive to pluralism and conflicts. If we get a clear picture of the principles that give shape to conceptions of agonistic politics, we can make institutions subject of an investigation that is guided by the question of whether the features of the institution in question are conducive or obstructive to a politics driven by these principles. In case they prove conducive, the institution can be valued as an institution which notwithstanding the question of whether it constitutes a necessary component of an ideally realised agonistic democracy promises to contribute to the enhancement of democracy s agonistic character. And in case the institution s features prove obstructive to agonism, the institution should be either dismissed or revised in ways that realise its hitherto hidden potential to function agonistically. If the latter applies, the reasons that have led to an assessment of the established features as problematic serve as a basis from which proposals for reform can be developed. Because once one has identified particular design features that prove problematic for agonism, one has the requisites to develop alternative features that have the capacity to turn the institution into one that, in contrast to its original version, is able to render democracy s institutional setting more conducive to agonism. Even though there may be kinds of institutions that prove obstructive to agonism altogether, one can suspect that in most cases prospects for agonistic reforms will open up. Depending on the ways in which they are set up and relate to others, institutions can function in very different ways. When we refer to particular institutions like, for instance, the parliament or the constitution, what we generally have in mind is rough ideas about political functions and competences rather than institutions already fleshed out in detail. In pursuing the project of rendering visible their approaches institutional implications, agonists should scrutinise the range of options for institutional design in particular cases, point out where they suggest different consequences than liberal and deliberative theories and thus sketch the steps democracies would be well advised to take in order to attune their institutional settings to the challenges resulting from deep dissent. One crucial challenge this approach implies is the task to choose which institutions to deal with. Whereas there might be other useful criteria available, I suggest to connect to the debate in which agonists have proven their theories value the debate on pluralism. Political liberals and deliberative theorists have been engaging with institutions they consider to be particularly suited to the conditions of pluralism and developed suggestions which, due to their theories persistent dominance in political theory, influence the wider perception of adequate institutional responses to pluralism to a significant extent. In order to substantiate agonistic democracy s status as a strand of political theory that has something meaningful to contribute to the debate on democratic responses to pluralism, a fruitful starting point for an application of the sketched approach would be to address institutions political liberals and deliberative theorists consider 8

9 particularly effective. By applying principles of agonistic politics, we are able to follow the steps outlined above, identify design features that prove problematic from an agonistic point of view and develop proposals for reform that are capable to enhance the agonistic character of the institution in question. In section 4, I will illustrate this approach by addressing one institution that has been receiving increasing interest by deliberative theorists: mini-publics. According to their proponents, minipublics are valuable democratic innovations capable to enhance democracy s responsiveness to the diversity of viewpoints and moral claims among the citizenry. In assuming that agonists and proponents of mini-publics share an inclination towards democratic innovations, I consider this institution to be an interesting test case for the sketched approach. But before turning to the test case it is necessary to deal with conceptions of agonistic politics in more detail. 3 Types and principles of agonistic politics In this section, I focus on the theories of agonistic democracy developed by Mouffe, Connolly and Tully. In acknowledging that these theories complex lines of argumentation cannot be appreciated comprehensively here, my aim is to expose the main features of their respective visions of agonistic politics. I will demonstrate that each implies specific ideas of what realising democracy under conditions of deep dissent means and, relatedly, how conflicts resulting from dissent should be dealt with. Mouffe s agonistic pluralism builds on the idea that political conflicts result from a moment of antagonism built into social relations. In drawing on her work co-authored with Ernesto Laclau, Mouffe distinguishes between politics, the ensemble of practices discourses and institutions which seek to establish a certain order and organize human coexistence (Mouffe 2000: 101) and the political, the dimension of antagonism that is inherent in human relations (ibid.). In order to grasp the nature of the political, it is necessary to consider the concept of a chain of equivalence (Laclau/Mouffe 2001: 127) which Laclau und Mouffe introduce to describe how collective identities come into being. Its basic idea is that identities lack any essence what constitutes them is differentiation from other identities. Consequently, identities only become united into one collective identity through differentiating from a shared other. Without becoming identical, its constituent parts are equivalent in the sense that they find themselves in a coalition against something they jointly oppose (Laclau/Mouffe 2001: 128). And the fact that all collective identities consist in such chains of equivalence relying on coalitions against others accounts for why a dimension of antagonism is inherent to social relations. For politics, this means that there can be no relation without this antagonistic element effecting that parties conceive of each other as opponents. Politics [ ] is always concerned with the creation of an us by the 9

10 determination of a them. (Mouffe 2000: 101) Whereas these processes explain why conflicts are inevitable, the crucial question they raise is how democracy can be realised given enduring conditions of mutual rejection. In a nutshell, Mouffe sees the key to this challenge, first, in accepting the impossibility to overcome confrontations and, second, in making confrontations cast off their potentially violent character. What liberal democratic politics requires is that the others are not seen as enemies to be destroyed, but as adversaries whose ideas might be fought, even fiercely, but whose right to defend those ideas is not to be questioned. To put it another way, what is important is that conflict does not take the form of an antagonism (struggle between enemies) but the form of an agonism (struggle between adversaries). (Mouffe 2013: 7) This transformation of antagonisms into agonisms is enabled, Mouffes argues, through shared adhesion to the political principles of a liberal democratic regime (Mouffe 2005: 52f). Liberty and equality provide for the common ground political actors are in need of to conceive of each other not as dangerous but friendly enemies and, consequently, to abstain from fighting each other with violence. A necessary precondition for liberty and equality to fulfil this function is that there are many possible interpretations of those principles (Mouffe 2005: 52). They constitute the reference points of a conflictual consensus (Mouffe 2000: 103) which does not rest on a consensus regarding these principles content but in line with the notion of equivalences regarding positions that oppose liberty and equality. Within the bounds of what the democratic coalition allows for, parties should fight for their own interpretation of the principles to become hegemonic (Mouffe 2013: 7) as it is the ongoing confrontation among democratic alternatives in the course of which none can establish its position on a long term basis that keeps alive democracy s uniting promise that the required consensus is truly conflictual (ibid.). For Connolly too, identity is neither an individual phenomenon nor something that social actors possess in the sense of essential characteristics. Identity is relational and collective (Connolly 2002: xiv) because it signifies a position within social relations that is defined through ongoing processes of differentiation from other identities. My personal identity is defined through the collective constituencies with which I identify or am identified by others (as white, male American, a sports fan, and so on); it is thus further specified by comparison to a variety of things I am not. Identity, then, is always connected to a series of differences that help it be what it is. (Ibid.) In contrast to Mouffe, though, Connolly does not conclude from this constitutive relationship and the potential for conflict it implies that democratic politics cannot be but a tamed struggle for hegemony. From Connolly s point of view, the crucial challenge is to engage with a tension between pluralism and pluralisation the range of identities, relationships and rules that exists at a given time and processes that challenge the existing configuration to bring new things into 10

11 being (Connolly 2004: xiv). Acknowledging the pluralist character of the social, according to Connolly, requires acknowledging both dimensions because the plurality that exists beyond an established plurality is nothing but the latter s enabling condition. But such acknowledgment, Connolly argues, is hindered by the predominance of a culture of pluralism which focuses on the existing and misrecognises new possibilities of diversification by freezing moral standards of judgment condensed from past political struggles (ibid.). What a pluralist account of politics thus has to achieve is to point out how this culture can be shaken up so that challenges against existing constellations are recognised as necessary implications of pluralism. The reason why this is hard to achieve is that the resistance against processes of pluralisation is nourished by deeply entrenched socio-psychological dispositions. Connolly claims that actors are driven by a desire to realise their respective identity as wholeness and thus conceive of unforeseen shifts unsettling the status quo as threatening. As long as the ideal of identity as wholeness persists, actors react to such occurrences with resentment against the other whom they blame for the unsettling experience and punish by fending off its attempts for modifications (Connolly 2010: 226ff). Against the background of these considerations, it becomes clear why a pluralist politics requires more than a taming of conflicts; it requires an overcoming of the tendency to interpret new identities as threats. And this overcoming can only be achieved, Connolly argues, by a readiness to work tactically upon established constituency reactions (Connolly 2010: 227). This means actors are required to engage in critical, self-reflective processes in which they detach themselves from presumptions entrenched in their identities, affirm the contestable character of their perspectives and, consequently, acquire an openness towards pluralism and the uncertainty it implies (Connolly 2004: 36). The goal of such processes is that actors translate their newly gained openness into a political conduct driven by respect for opponents who possess roughly the same power and by a generous responsiveness towards opponents who challenge the existing order from weaker positions (Connolly 2005: 123ff). If a significant number of actors comes to cultivate these virtues, political conflicts can be dealt with in a way that adequately recognises the tension between pluralism and pluralisation through furthering the latter and thus constantly re-establishing the limits of the former. Tully s account of agonism differs from the other two in that it explicitly affirms the notion of deliberation. Whereas it is therefore often classified as an account of deliberative democracy (e.g., Chambers 2003: 310), Tully in fact develops a concept of dialogue that dismisses any trust in argumentation s capacity to identify rational reasons and, instead, affirms the idea of reasonable disagreement all the way down (Tully 2002: 218). Even though he refers to the term deliberation, the nature of the process through which, according to Tully, groups should deal with their conflicts is more adequately characterised as dialogical negotiation. Tully values dialogues as indispensable for democracy because they represent the only fairly reliable and effective way to work up a norm of mutual recognition that does justice to the diversity and 11

12 changeability of the member of contemporary political associations (Owen/Tully 2007: 283). Through participation in dialogues, citizens make use of their democratic right to have a say over the rules under which they live and make sure that the rules can prove acceptable to all (Owen/Tully 2007: 280ff). Whereas these abstract ideas about the need for dialogues are in line with what deliberative democrats expect deliberations to achieve, Tully opts for realising them in a way that clearly diverges from the deliberative view. The difference finds expression in Tully s understanding of dialogues as negotiating practices (Tully 2014: 36, 48). On the one hand, participants are required to explain their own perspectives to those whom they disagree with and, reciprocally, to listen to what the others tell and explain to them (Tully 2007: 25f). The importance Tully attaches to a give and take of reasons becomes obvious by his calling audi alteram partem, always listen to the other side the first and perhaps only universalizable principle of democratic deliberation (ibid.). The status of the dialogue as negotiation, on the other hand, comes into view once one takes a more detailed look at the features of the process through which audi alteram partem is realised. First, dialogues are the exercise of practical, not theoretical reason (Tully 2008: 163) which means that the range of legitimate contributions encompasses a multiplicity of modes of expressions like, for instance, [ ] presenting a reason, a story, an example, a comparison, a gesture or a parable for consideration, showing rather than saying, expressing disagreement, deferring or challenging, [ ] stonewalling, feet-dragging and feigning, dissenting through silence, breaking off talks [ ] and countless other discursive and nondiscursive activities with [sic!] make up deliberative language games. (Tully 2002: 223) Dialogue thus include modes of interaction that go beyond argumentation. Second, dialogues are agonistic which means that no agreement reached in the course of the interchange is expressive of a consensus. Any agreement is partial or conditional to some extent; open to possible redescription and challenge. (Tully 2008: 163) In the face of these characterisations, it becomes clear that the kind of agreements Tully s dialogue is targeted at are compromises, decisions that combine elements of all those perspectives involved in a conflict and, due to this composite character, prove acceptable to all participants despite their persisting dissent. Thus, what citizens have to do when participating in dialogue is not deliberating in the strict sense of the word but negotiating in a manner whose moral quality results from the requirements any solution must realise to be acceptable to deeply dissenting parties. The brief overviews over the three agonistic theories show that each is expressive of distinct ideas about adequate ways of dealing with conflicts. According to Mouffe, politics is a struggle for hegemony which, in order to be carried out democratically, requires a taming of conflicts. Connolly envisions politics as a process of pluralisation in the course of which conflicts are regulated in favour of new and formerly excluded identities. For Tully, politics is a dialogical 12

13 process directed towards mutually accepted terms of recognition which requires conflict parties to negotiate their conflicts and arrive at fair compromises. Whereas these principles constitute the normative ideas of what, according to the agonists, politics has to achieve under conditions of pluralism, the respectively suggested notions of ethos provide for the conditions under which processes driven by these principles can be realised. Liberty and equality constitute the shared ground opponents are in need of to tame their conflicts; the pluralist ethos acquired through critical self-reflection effects a readiness to accommodate pluralising projects; a commitment to audi alteram partem enables the definition of fair compromises. Thus, agonistic conceptions of political ethos are elements of more comprehensive arguments whose nature is captured by the above depicted principles. 4 Agonism and democratic innovations: the case of mini-publics In what follows, I illustrate the approach I suggested in the first part of the paper by applying the principles of agonistic politics to institutions that have been attracting increasing attention by theorists dealing with democratic innovations, namely mini-publics. The term mini-public is generally applied to a class of institutions that directly engage citizens, promote democratic deliberation and have at times been institutionalized into contemporary decision making processes (Ryan/Smith 2014). Whereas proponents of mini-publics disagree about the details of a definition and, accordingly, about the boundaries of the category (ibid., Dryzek/Goodin 2006, Fishkin 2009, Fung 2003, Smith 2009), it is usually assumed that mini-publics are issuespecific and timely limited and gather a number of citizens that ranges from 12 to 500 (Smith 2009: 73ff). Why is it worthwhile to choose mini-publics as an example for an application of the different visions of agonistic politics? First, mini-publics fulfil the criterion defined in section 2.2. Deliberative democrats value mini-publics because they consider them particularly suited to map the plurality of viewpoints among the citizenry and make it a relevant factor for political decision-making. Expression of this trust in mini-publics capacity to enhance responsiveness towards pluralism is the fact that many of the most widely known practical examples were implemented to address contested moral issues. This applies in the cases of the Citizens Council established by the National Institute for Clinical Excellence in the UK to deal with questions related to allocation in health care (Gutmann/Thompson 2004: 97f) and the Danish Consensus Conferences set up to debate new scientific and technological developments which raise serious social and ethical concerns (Smith 2009: 77). That means the idea that mini-publics are particularly valuable institutions for pluralist democracies is not only spread in academia but also influences recent developments in political practice. Against the background of my suggestion to apply agonism to institutions that are currently considered effective means to address the challenge of pluralism, mini-publics thus present themselves as a fruitful starting 13

14 point. Second, it can be assumed that agonists share with proponents of mini-publics an inclination towards democratic innovations. All agonists relate their criticism of public reasoncentred political theory to a critical analysis of the current state of affairs in liberal democracies and thus suggest that their views on politics demand modifications of both theory and practice. And as it is the very purpose of democratic innovations to modify the status quo, it seems adequate to put forward the hypothesis that the attempt to further democratic innovations falls on fertile ground with agonistic democracy. Building on these considerations, I will probe on whether and in what ways agonism changes our understanding of mini-publics value and the design measures required to realise it. In a first step, I take a closer look at deliberative democrats arguments for mini-publics and sketch the design measures they suggest. In a second step, I apply the agonistic principles as critical measures to evaluate these arguments and explore under which conditions mini-publics can be conducive to agonistic politics. 4.1 Deliberative mini-publics What makes deliberative democrats interested in innovating democracies institutional settings through implementing mini-publics? Graham Smith asserts that mini-publics have raised the attention of deliberative democrats because they appear to realize a compelling combination of two democratic goods: inclusiveness and considered judgement. (Smith 2012: 93) And the reasons why mini-publics are expected to realise these goods is that they, first, have some claim to representativeness of the public at large (Dryzek/Goodin 2006: 221) and, second, use active facilitation that encourages fairness in proceedings, in particular, support and encouragement for less confident and politically skilled participants to voice their perspective (Smith 2012: 94). It is thus their mode of selection as well as their procedural conditions for deliberation through which mini-publics, according to their proponents, present themselves as distinct and particularly valuable institutions. Mini-publics claim to representativeness rests on the application of a selection technique that realises a reproduction of the wider public on the small-scale level. Whereas some theorists argue that a mini-public should constitute an accurate microcosm of the public (Fishkin 2009: 55), others emphasise that representativeness does not require that social characteristics and viewpoints [ ] be present in the same proportions as in the larger population (Dryzek/Goodin 2006: 221) but that the diversity of social characteristics and plurality of initial points of view in the larger society are substantially present in the deliberating mini-public (ibid.). Disregarding the question for statistical accuracy, proponents of mini-publics agree that the most effective way to realise the representative character of mini-publics is the use of random 14

15 selection. Through gathering a group of citizens with the help of random selection techniques based on stratified samplings, designers safeguard that members of various social groups are recruited and prevent that numerically small groups are excluded from participation (Smith 2009: 81). But the use of random selection oriented towards a balanced representation of the public s socio-demographic characteristics does not only realise the constitution of a diverse body of citizens (Smith 2012: 94) and basic fairness considerations by making sure that no citizen has a better chance than others to become a participant (Smith 2012: 95). One further important effect is that it is lay citizens and non-partisans (Dryzek/Goodin 2006: 221) who are selected for participation. This effect is crucial as regards mini-publics capacity to realise the second democratic good: considered judgement. For deliberative democrats, the valuable implication of selecting non-partisans is that the design of mini-publics can be said to orientate citizens towards considerations of the public interest, rather than their own self-interest (Smith 2009: 94). The reason for this is that participants who are not selected to represent particular social groups or interests in the strong sense are typically more open to changing their views as they hear new evidence and insights (ibid.). But representativeness and non-partisanship are only necessary, not sufficient conditions for a deliberative process to generate considered judgements. In addition, designers are required to arrange for procedural conditions that help the process to unfold the criteria of high-quality deliberation (Fishkin 2009: 99), substantive balance, conscientious (rather than strategic) participation, diversity of viewpoints represented and equal consideration of viewpoints regardless of the status of those offering them in the discussion (ibid.). Among the design measures deliberative democrats consider most effective in facilitating inclusive deliberation are encounters between participating citizens and experts who provide the participants with information on the issue in question as well as small group discussions in which professionally skilled moderators protect the procedure s fairness by, for instance, preventing confident individuals from dominating the discussion and encouraging less confident ones to speak out (Smith 2009: 85f, 94). From this we can conclude that deliberative democrats value mini-publics due to their providing deliberative spaces which are protected from the inequalities inherent within civil society and political institutions (Smith/Wales 2000: 52) and thus enable citizens to arrive at informed and considered judgements on the issues in question. Although the theorists I referred to do not understand their approaches as applications of Habermas s and Rawls s theories, the rationale underlying the design measures they propose clearly reflects an endorsement of the basic idea guiding public reason-oriented political theory, namely that given accommodating conditions reason-giving enables the identification of reasonable solutions. Whilst the proposed design measures are meant to provide for these accommodating conditions on a micro-political level, deliberative theorists understanding of the macro-political functions mini-publics are meant to fulfil are likewise expressive of this trust in the quality of results generated in deliberation. Even 15

16 though the discussion on mini-publics macro-political uptake is only just emerging, there is a general tendency to expect mini-publics to improve the quality of political decision-making via two functions. First, mini-publics are considered valuable consultation bodies for policy makers (Dryzek/Goodin 2006); second, they are meant to distil, synthesise and shape information for public consumption (Niemeyer 2014). In line with these ideas, discussions that identify options to enhance mini-publics macro-political impact suggest to intensify forms of communication among mini-publics and both policy makers and the wider public (Bächtiger et al. 2014). 4.2 Agonistic mini-publics In applying principles of agonistic politics to mini-publics I will proceed in two steps. In the first, I make the specific design features which, according to deliberative democrats, make up the value of mini-publics subject of a critical evaluation. In the second step, I turn the focus to the macro-political functions presupposed by deliberative democrats and ask whether agonism suggests to assign different functions to mini-publics. Among the agonistic theories dealt with in this paper, Tully s is the only one explicitly stressing the political value of dialogical interchange. But even though Connolly and in particular Mouffe distance themselves more radically from the idea of politics as deliberation, this does not mean that there is no place for dialogues in their accounts of agonistic politics whatsoever. A process, for instance, in which antagonistic opponents identify positions they jointly reject can have the taming effect Mouffe argues for if it renders visible a shared ground they did not know about before. Likewise, a dialogue can be an effective part of a pluralising process as it is developed by Connolly. If it gives the new and formerly excluded groups an opportunity to articulate their demands and display aspects of the existing order that pose obstacles to their political projects, the requirements for successful processes of pluralisation become visible. Without suggesting that dialogues are necessary or even the only elements of processes that realise a taming of conflicts or further the demands of formerly excluded parties, dialogues can prove fruitful for agonism if they contribute to effects that are expressive of the principles carved out above. But the question is: Under which conditions do mini-publics provide spaces for dialogues that are likely to realise these possible agonistic effects? Let us first address the task of selecting participants and the argument for random selection. As has been shown, deliberative democrats value random selection techniques because they ensure that the body is representative of the society and constituted by non-partisans. Against the background of a trust in the rationalising force of reasoning, both aspects indeed present themselves as fruitful design features. If it is assumed that argumentation has the potential to identify the most convincing reasons, it is sensible to opt for design measures that realise the 16